Unboxing the Legend Continues: Amiga 1000 Mouse Restoration.

I recently received a boxed Amiga 1000 which was in excellent condition, but required a little work. It came with two original Amiga 1000 mice and the keyboard, all of which needed work. Today I’ll cover the mouse restoration.

Introduction

As I mentioned in the previous post, this Amiga 1000 was on-loan with an option to purchase. On the day I published that post, I went ahead and purchased the Amiga (I paid more than the original owner asked for it). I’m very happy to own it.

The Amiga 1000 mice may look like regular tank mice, and functionally they are. But what makes them rare and valuable is the connector to the Amiga. Instead of a regular horizontal joystick-like connector, they use a unique right-angle connector.

Mouse #1

So, the first mouse I’m going to look at is the one that I tried originally with the Amiga 1000, it had no horizontal movement. So, I disassembled it to see what state it was in.

It was really dirty, I started cleaning it up, with the hope that it was just dirt that was causing the failure. Once cleaned, I plugged it into my Amiga 500 test motherboard to try it. The horizontal movement still didn’t work, but I noticed something. A telltale odour of magic smoke. I couldn’t see where it was coming from, but the smell was definitely coming from the mouse itself.

I turned the machine off, booted a thermal camera and tried again to see if I could see something burning up.

Nothing (it is a warm day in my workshop). Also, the smell has gone. I figured that whatever was dying has now died. The mouse is basically made up of LEDs, photo diodes and a comparator (HA17339). I ran a diode test on the LEDs and photo diodes. They all passed. So, my guess was that the comparator failed. The HA17339 is no longer in production, but it is basically an LM339, which is very easy to obtain, I have dozens of new/old stock LM339s from an Acorn repair centre. I decided to replace this.

This didn’t solve it. For the next step, I need to talk about how the mouse works. I also realised during the repair that the smell was just the very old fluxy PCB getting warm.

Further diagnosis

Each axis has infra-red LEDs and two phototransistors. The ball spins a wheel on each axis which has holes in it. Due to the position of the phototransistors, they don’t quite turn on at the same time. So, the order in which they turn on and off determines direction. A great image explaining this can be found on this Apple Mouse description.

Now, the photo transistors won’t produce TTL levels. The comparator compares the voltage produced with a known constant, and uses that to produce a TTL high or low output. The circuit diagram can be seen here.

When checking the outputs, I noticed that pin 13 never changed. That phototransistor was likely not producing a high enough output. I could replace them, but I’d likely need to do the calibration step I did next anyway because it is very likely the model used is no longer in production.

You can see each comparator in the diagram uses a voltage divider for the negative input. It is hard to see the numbers, but they are 6.8K and 270ohm. This makes the high level 0.191v, assuming a 5v input. I noticed that this phototransistor was not quite getting that high. I replaced the 270ohm with a 150ohm to make the detection level 0.108v.

In this photo, the 270ohm is the left of the two chip resistors above the bottom encoder wheel. Yes, this 1985 mouse used SMD resistors!

This solved the axis read problem. But there was another problem.

My dumb mistake

When trying to figure out the issue with the phototransistors, I removed the encoder wheel. When doing so, I accidentally broke the peg on the optical array side that holds the axel in. I quickly designed and 3D printed a replacement.

I reassembled the mouse and gave it a final test.

Mouse #2

This mouse had some physical issues, the cable had polystyrene from the box melted onto the cable and one of the bottom skates is missing. That being said, I plugged it into my test Amiga 500 and it worked perfectly.

Luckily, it isn’t as dirty inside. The contact rollers needed cleaning, and I cleaned the ball. I then scrubbed the cable with isopropyl alcohol to remove the melted polystyrene.

I removed the other skate at the back end of the mouse so that I could measure the depth of it and so that I can make sure both at the back will be the same depth. Which should stop any potential wobble on the table. The originals are around 11.4mm by 1mm thick. I managed to find 10x1mm PTFE disks on eBay which should work perfectly. The section with the missing pads was cleaned using isopropyl to get rid of dirt and remaining glue. After that, I stuck the replacements on using double-sided tape. I could have used glue, but I figured this was easier if they needed to be removed in the future.

With that, I have two working Amiga 1000 mice. Ideal for a two-player Lemmings session. Next up will be the keyboard repair!